What to do with meter? The question for poets is simple: employ it, reinvent it, or leave it alone. For critics, the question is answered with greater difficulty, though critics might be said to fall into three corresponding camps: employing it (by relating it to the poem’s subject matter, or investing it with political and cultural significance), reinvent it (devising new schemes and notations for describing and measuring its variations), or leave it alone, recognizing that it is essential to poetry but fundamentally resistant to critical analysis.

I tend to fall into the third category, finding that though the relation of rhythm and meter to subject matter might be valuable (the abruptly shifting time-signatures in Wordsworth’s “Ode” are an especially compelling example), it does not require the full resources of scansion to achieve insights; and, more aggressively, finding that new schemes for describing meter to accommodate the varieties of stresses and half-stresses are not only of limited critical use, but are detrimental to the sense of what meter is.

Though it’s not a full-formed thought, my sense is that a critical payoff of metrical analysis will come from a rich description of what meter is. The basic account is pretty clear: meter is the abstract and ideal pattern of stressed and unstressed syllables in a poem. But add to this: meter is an ideal pattern that can never be perfectly achieved owing to the natural variation of stress and rhythm in the language, and which therefore is an attempt at selecting, isolating, and elevating one pattern of stresses and rhythms from all of the variations that exist in a given phrase or unit of language. “This syllable could be heard as stressed or unstressed; for the sake of the meter, please hear it as stressed”–often that is a choice, an indeterminacy, that meter asks us to resolve. Poetry is what happens when the variations of the voice aspire to the ideal of meter–or what happens when the ideal of meter works on the variations of the voice. Poets who are masters of meter know what can and what cannot be asked of the voice; but they also know when, in a  given poem, the demands should be strongest and when they should be weakest.

When I say that I find it detrimental to devise new schemes for marking half- or third-stressed syllables, I do not mean that we should not notice these when we read; but it is not right to pretend that these are an account of meter. Meter is that abstract pattern of binaries, stressed or unstressed; it may be that the pattern includes intentional indeterminacies, but the fact of the indeterminacy matters; they should not be pinned down by a more nuanced system of measurement.

It might be useful to notice, beneath a binary account of meter, which syllables are being asked to strain upward the most, or which are being forced downward into unstressed positions; though a poem would fail if the pattern of stressed or unstressed syllables were imposed from without, the sense that the poet is isolating what is in effect a low-frequency stress among the range of possible stresses, or selecting a minority among the variations, is essential to appreciate meter. They are often doing so rather than yield to a substitution; they are insisting on meter, even though it means asking a reader to draw on a phonic possibility that is not usually realized when speaking a word. It is in the realm of possibility, but rare; it is asking the reader to distill the language.

But to say that it is useful to notice which syllables are being forced up or down the most is not the same as offering an account of meter that denies the ideal pattern to which they aspire.

Robert Browning, for instance, should not be reduced to marks of half-stressed and third-stressed syllables; the real problem when reading his poetry is deciding whether he is asking for us to hear something subtle within the range of a word’s or syllable’s possible stresses so that the word or syllable stretches to conform to a meter–or whether he is asking us to notice a substitution.

The difference feels especially relevant for a poet who writes dramatic monologues: are we being asked to hear the voice operating at a peculiar pitch, aspiring to the meter against the odds, or are we to hear the voice abandoning the regularity of meter altogether–or do neither happen, and does the voice happily, naturally and happily, find a meter? The three choices feel directly relevant to the emotional life of the speakers, and possibly to the judgments that Browning’s poetry implicitly passes on them.

Browning, with greater frequency than any other mid-Victorian, asks the reader to decide between the options (uneasy aspiration, substitution, happy concord); but Tennyson, because of the consistency of his metrical happiness and ease, is more remarkable when he demands deliberation between uneasy aspiration and substitution; when he asks either that the reader to find a stress that is plausible but only barely, or else ponder a substitution.

I suspect that he does so most in Maud, a poem that sets the versatility of lyrical forms against a fragmenting mind; the potential of acoustic harmony on the surge of psychological discord. It is a poem that aspires to an ideal, and that is compromised in the situation of its aspiration: a description of both its meter and its subject.